Changjun Jiang’s research while affiliated with Tongji University and other places

Financial time series (FinTS) record the behavior of human-brain-augmented decision-making, capturing valuable historical information that can be leveraged for profitable investment strategies. Not surprisingly, this area has attracted considerable attention from researchers, who have proposed a wide range of methods based on various backbones. However, the evaluation of the area often exhibits three systemic limitations: 1. Failure to account for the full spectrum of stock movement patterns observed in dynamic financial markets. (Diversity Gap), 2. The absence of unified assessment protocols undermines the validity of cross-study performance comparisons. (Standardization Deficit), and 3. Neglect of critical market structure factors, resulting in inflated performance metrics that lack practical applicability. (Real-World Mismatch). Addressing these limitations, we propose FinTSB, a comprehensive and practical benchmark for financial time series forecasting (FinTSF). To increase the variety, we categorize movement patterns into four specific parts, tokenize and pre-process the data, and assess the data quality based on some sequence characteristics. To eliminate biases due to different evaluation settings, we standardize the metrics across three dimensions and build a user-friendly, lightweight pipeline incorporating methods from various backbones. To accurately simulate real-world trading scenarios and facilitate practical implementation, we extensively model various regulatory constraints, including transaction fees, among others. Finally, we conduct extensive experiments on FinTSB, highlighting key insights to guide model selection under varying market conditions. Overall, FinTSB provides researchers with a novel and comprehensive platform for improving and evaluating FinTSF methods. The code is available at https://github.com/TongjiFinLab/FinTSBenchmark.

After decades of evolution, the financial system has increasingly deviated from an idealized framework based on theorems. It necessitates accurate projections of complex market dynamics and human behavioral patterns. With the development of data science and machine intelligence, researchers are trying to digitalize and automate market prediction. However, existing methodologies struggle to represent the diversity of individuals and are regardless of the domino effects of interactions on market dynamics, leading to the poor performance facing abnormal market conditions where non-quantitative information dominates the market. To alleviate these disadvantages requires the introduction of knowledge about how non-quantitative information, like news and policy, affects market dynamics. This study investigates overcoming these challenges through rehearsing potential market trends based on the financial large language model agents whose behaviors are aligned with their cognition and analyses in markets. We propose a hierarchical knowledge architecture for financial large language model agents, integrating fine-tuned language models and specialized generators optimized for trading scenarios. For financial market, we develop an advanced interactive behavioral simulation system that enables users to configure agents and automate market simulations. In this work, we take commodity futures as an example to research the effectiveness of our methodologies. Our real-world case simulation succeeds in rehearsing abnormal market dynamics under geopolitical events and reaches an average accuracy of 3.4% across various points in time after the event on predicting futures price. Experimental results demonstrate our method effectively leverages diverse information to simulate behaviors and their impact on market dynamics through systematic interaction.

The landscape of financial transactions has grown increasingly complex due to the expansion of global economic integration and advancements in information technology. This complexity poses greater challenges in detecting and managing financial fraud. This review explores the role of Graph Neural Networks (GNNs) in addressing these challenges by proposing a unified framework that categorizes existing GNN methodologies applied to financial fraud detection. Specifically, by examining a series of detailed research questions, this review delves into the suitability of GNNs for financial fraud detection, their deployment in real-world scenarios, and the design considerations that enhance their effectiveness. This review reveals that GNNs are exceptionally adept at capturing complex relational patterns and dynamics within financial networks, significantly outperforming traditional fraud detection methods. Unlike previous surveys that often overlook the specific potentials of GNNs or address them only superficially, our review provides a comprehensive, structured analysis, distinctly focusing on the multifaceted applications and deployments of GNNs in financial fraud detection. This review not only highlights the potential of GNNs to improve fraud detection mechanisms but also identifies current gaps and outlines future research directions to enhance their deployment in financial systems. Through a structured review of over 100 studies, this review paper contributes to the understanding of GNN applications in financial fraud detection, offering insights into their adaptability and potential integration strategies.

Anti-money laundering (AML) is crucial to maintaining national financial security. Contemporary AML methods focus on homogeneous mining or unitary money laundering pattern. These methods ignore a characteristic of gang operation in money laundering. Thus, in this paper, we propose a multi-view graph based hierarchical representation learning method, named MG-HRL, to mine organized money laundering groups. In particular, we extract multi-level representations of transaction subgraphs, including transaction features, user features, structural features, and high-order association features from multiple observational perspectives. To learn the correlation between users, we model transaction networks as heterogeneous information networks (HINs) and design six meta-paths related to money laundering scenarios to mine correlations among users. Combining with correlation representations of users, we propose a heterogeneous hypergraph representation learning method to learn high-order representations of transaction subgraphs. Through hierarchical representation learning, the MG-HRL achieves full exploration of money laundering groups. Finally, we conduct experiments on two public transaction datasets. The result shows that MG-HRL method performs better than other state-of-the-art baselines.

Accurately assessing and forecasting bank credit ratings at an early stage is vitally important for a healthy financial environment and sustainable economic development. However, the evaluation process faces challenges due to individual attacks on the rating model. Some participants may provide manipulated information in an attempt to undermine the rating model and secure higher scores, further complicating the evaluation process. Therefore, we propose a novel approach called the preferential selective-aware graph neural network (PSAGNN) model to simultaneously defend against feature and structural nontarget poisoning attacks on Interbank credit ratings. In particular, the model establishes a phased optimization approach combined with biased perturbation and explores the Interbank preferences and scale-free nature of networks, to adaptively prioritize the poisoning training data and simulate a clean graph. Finally, we apply a weighted penalty on the opposition function to optimize the model so that the model can distinguish between attackers. Extensive experiments on our newly collected Interbank quarter dataset and case studies demonstrate the superior performance of our proposed approach in preventing credit rating attacks compared to state-of-the-art baselines.

This paper is concerned with the problem of distributed Kalman filtering over sensor networks under two-bitrate periodic coding strategies. Initially, the optimal estimates for sensor individuals are acquired using the conventional Kalman filter. Subsequently, the information pair, consisting of the local estimate and the corresponding covariance, is exchanged among their immediate neighbors to achieve cooperative estimation. Due to the constrained network bandwidth, a vector/matrix quantization approach is formulated to quantize the information pair. The output of this quantization establishes a conservative bound for the actual covariance. A two-bitrate periodic coding strategy is proposed, where the encoded bits of the quantizer outputs are divided into two separate parts, namely the most significant and least significant bits, following a periodic transmission principle. It is demonstrated that the estimation preserves a consistency property over the sensor networks as the reported error covariance always serves as an upper bound for the actual error covariance. It is shown that the mean-square estimation errors are bounded when certain conditions regarding collective observability and network connectivity are satisfied. Finally, the effectiveness of the proposed algorithm is verified through a numerical example.

Heterogeneous data silos hinder the application of deep learning in the Internet of Things. As a dealing scheme, personalized federated learning (pFL) distributedly customizes multiple local models for these silos. Most pFL methods directly use a global model to assist local model optimization, ignoring the performance drop caused by irrelevant or misleading global-model knowledge. To address this, we propose an adaptive knowledge recomposition approach (FedAKR), which refines relevant and correctly-leading knowledge from the global model and recomposes it into the local model to promote personalization. Specifically, FedAKR provides a discrete wavelet transform-based method to recompose different kinds of knowledge in the same representation space. Facilitated by this common space, we introduce an enriched local optimization objective to establish a causal relationship between refined global-model knowledge and recomposed local-model knowledge. The relationship guides effective and efficient knowledge refinement, thereby promoting personalization. Besides, we provide the theoretical proof of convergence for our novel pFL approach. Extensive experiments demonstrate that FedAKR achieves interpretable improvements, higher performance over twelve state-of-the-art methods, and the potential to further integrate pre-trained large models.

Each activity in the service workflow interacts with services as required to meet complex business needs and quickly adapt to market changes. The design of each activity's input/output interface parameters influences whether it can successfully map to appropriate and interactive services. In practice, suitable activity interface parameters should possess 3-features: realism, relevance, and compatibility, as popular parameters originating from the real world and closely related to activity semantics are apt to match user-expected services. However, existing research requires expert specification or ontology-based inference, resulting in outdated, inconsistent parameters that lack necessary elements, making it challenging to match expected services. Therefore, we propose an automated method combining Transformer and weighted HITS to recommend interface parameters with 3-features on activity function requirement. It filters similar Endpoints (EPs) based on the activity's semantics by supervised Transformer-based learning of multidomain APIs and unsupervised EPs matching. Next, a nodeweighted heterogeneous graph is built based on similar EPs and their interface parameter relationships. We then apply a nodeweighted HITS to explore mutual gain relationships within the graph and calculate parameter compatibilities. Finally, a topk non-redundant compatible parameter list and corresponding different formats are recommended for the activity. The method's effectiveness and efficiency are verified using a real API service dataset from RapidAPI

The landscape of financial transactions has grown increasingly complex due to the expansion of global economic integration and advancements in information technology. This complexity poses greater challenges in detecting and managing financial fraud. This review explores the role of Graph Neural Networks (GNNs) in addressing these challenges by proposing a unified framework that categorizes existing GNN methodologies applied to financial fraud detection. Specifically, by examining a series of detailed research questions, this review delves into the suitability of GNNs for financial fraud detection, their deployment in real-world scenarios, and the design considerations that enhance their effectiveness. This review reveals that GNNs are exceptionally adept at capturing complex relational patterns and dynamics within financial networks, significantly outperforming traditional fraud detection methods. Unlike previous surveys that often overlook the specific potentials of GNNs or address them only superficially, our review provides a comprehensive, structured analysis, distinctly focusing on the multifaceted applications and deployments of GNNs in financial fraud detection. This review not only highlights the potential of GNNs to improve fraud detection mechanisms but also identifies current gaps and outlines future research directions to enhance their deployment in financial systems. Through a structured review of over 100 studies, this review paper contributes to the understanding of GNN applications in financial fraud detection, offering insights into their adaptability and potential integration strategies.